flint ROCmFPX needs about 24.6 GB of memory at Q4_0 (a 23.3 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 4 quants, from 22.1 GB (Q4_0) to 41.1 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 3.3 GB.
flint ROCmFPX is a 39.8B-parameter Flint model in GGUF format, quantized by Harsha001. Its trained context length is 256K tokens. The repository has 4 quant files, from 22.1 GB (Q4_0) to 41.1 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 22.1 GB | 23.4 GB | Too big | Get file → | |
| Q4_0 | 23.3 GB | 24.6 GB | Too big | Get file → | |
| Q6_K | 32.9 GB | 34.2 GB | Too big | Get file → | |
| Q8_0 | 41.1 GB | 42.4 GB | Too big | Get file → |
What GPU do you need for flint ROCmFPX?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q4_0 (22.1 GB file, about 23.4 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_0 (23.3 GB file, about 24.6 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (41.1 GB file, about 42.4 GB in use).
Which flint ROCmFPX quant should you choose?
Q4_0 (23.3 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is Q4_0 at 22.1 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about flint ROCmFPX
How much VRAM does flint ROCmFPX need?
About 24.6 GB at Q4_0 with an 8K context window. That is the 23.3 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run flint ROCmFPX on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 23.4 GB, so you would need a larger GPU or CPU offloading.
What is the file size of flint ROCmFPX GGUF?
It depends on the quant: Q4_0 is 22.1 GB, Q4_0 is 23.3 GB, Q6_K is 32.9 GB, Q8_0 is 41.1 GB.
What context length does flint ROCmFPX support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.3 GB at 32K.
Where can I download flint ROCmFPX GGUF?
The files are on Hugging Face in the Harsha001/flint-ROCmFPX repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does flint ROCmFPX use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.